UNFOLD-SENSE: A parallel MRI method with self-calibration and artifact suppression

Self-supervised learning of physics-guided reconstruction neural networks without fully sampled reference data

PURPOSE

To develop a strategy for training a physics-guided MRI reconstruction neural network without a database of fully sampled data sets.

METHODS

Self-supervised learning via data undersampling (SSDU) for physics-guided deep learning reconstruction partitions available measurements into two disjoint sets, one of which is used in the data consistency (DC) units in the unrolled network and the other is used to define the loss for training. The proposed training without fully sampled data is compared with fully supervised training with ground-truth data, as well as conventional compressed-sensing and parallel imaging methods using the publicly available fastMRI knee database. The same physics-guided neural network is used for both proposed SSDU and supervised training. The SSDU training is also applied to prospectively two-fold accelerated high-resolution brain data sets at different acceleration rates, and compared with parallel imaging.

RESULTS

Results on five different knee sequences at an acceleration rate of 4 shows that the proposed self-supervised approach performs closely with supervised learning, while significantly outperforming conventional compressed-sensing and parallel imaging, as characterized by quantitative metrics and a clinical reader study. The results on prospectively subsampled brain data sets, in which supervised learning cannot be used due to lack of ground-truth reference, show that the proposed self-supervised approach successfully performs reconstruction at high acceleration rates (4, 6, and 8). Image readings indicate improved visual reconstruction quality with the proposed approach compared with parallel imaging at acquisition acceleration.

CONCLUSION

The proposed SSDU approach allows training of physics-guided deep learning MRI reconstruction without fully sampled data, while achieving comparable results with supervised deep learning MRI trained on fully sampled data.

Non-Cartesian GRAPPA and coil combination using interleaved calibration data - application to concentric-ring MRSI of the human brain at 7T

PURPOSE

Proton MR spectroscopic imaging (MRSI) benefits from B₀ ≥ 7T and multichannel receive coils, promising substantial resolution improvements. However, MRSI acquisition with high spatial resolution requires efficient acceleration and coil combination. To speed up the already-fast sampling via concentric rings, we implemented additional, non-Cartesian, hybrid through-time/through-k-space (tt/tk)-generalized autocalibrating partially parallel acquisition (GRAPPA). A new multipurpose interleaved calibration scan (interleaved MUSICAL) acquires reference data for both coil combination and PI. This renders the reconstruction process (especially PI) less sensitive to instabilities.

METHODS

Six healthy volunteers were scanned at 7T. Three calibration datasets for coil combination and PI were recorded: a) iMUSICAL, b) static MUSICAL as prescan, c) moved MUSICAL as prescan with misaligned head position. The coil combination performance, including motion sensitivity, of iMUSICAL was compared to MUSICAL for single-slice free induction decay (FID)-MRSI. Through-time/through-k-space-GRAPPA with constant/variable-density undersampling was evaluated on the same data, comparing the three calibration datasets. Additionally, the proposed method was successfully applied to 3D whole-brain FID-MRSI.

RESULTS

Using iMUSICAL for coil combination yielded the highest signal-to-noise ratio (SNR) (+9%) and lowest Cramer-Rao lower bounds (CRLBs) (-6%) compared to both MUSICAL approaches, with similar metabolic map quality. Also, excellent mean g-factors of 1.07 and low residual lipid aliasing were obtained when using iMUSICAL as calibration data for two-fold, variable-density undersampling, while significantly degraded metabolic maps were obtained using the misaligned MUSICAL calibration data.

CONCLUSION

Through-time/through-k-space-GRAPPA can accelerate already time-efficient non-Cartesian spatial-spectral 2D/3D-MRSI encoding even further. Particularly promising results have been achieved using iMUSICAL as a robust, interleaved multipurpose calibration for MRSI reconstruction, without extra calibration prescan.

Dynamic field-of-view imaging to increase temporal resolution in the early phase of contrast media uptake in breast DCE-MRI: A feasibility study

PURPOSE

To increase diagnostic accuracy of breast MRI by increasing temporal resolution and more accurately sampling the early kinetics of contrast media uptake. We tested the feasibility of accelerating bilateral breast DCE-MRI by reducing the FOV, allowing aliasing, and unfolding the resulting images.

METHODS

Previous experience with an "ultrafast" protocol for bilateral breast DCE-MRI (6-10 s temporal resolution) showed that the number of significantly enhancing voxels is very low in the first 30-45 s after contrast media injection. This suggests that overlap of enhancing voxels in aliased images will be very infrequent. Therefore, aliased images can be acquired during the first 30-45 s after contrast media injection and unfolded to produce full-FOV images with few errors. In a proof-of-principle test, aliased images were simulated from the first 30 s of full-FOV acquisitions. Cases with relatively dense early enhancement were selected to test this method in a worst-case scenario. In an initial test, an FOV of 60% the size of the full FOV was simulated. To reduce the probability of errors due to overlapping voxels in aliased images, we then tested a dynamic FOV approach. The FOV was progressively increased so that enhancing voxels could not overlap at multiple time-points, and areas where enhancing voxels overlapped at a given time-point could be unfolded by interpolating between the preceding and subsequent time-points (acquired with different FOVs). The simulated FOV sizes for each of the time-points were 31%, 44%, and 77% of the full FOV. Subtraction images (post- minus precontrast) were generated for aliased images and filtered to select significantly enhancing voxels. Comparison of early, highly aliased images, with later, less aliased images then helped to identify the true locations of enhancing voxels.

RESULTS

In the initial aliasing simulations, an average of 2.9% of the enhancing voxels above the chest wall overlapped in the aliased images (range 0.1%-6.7%). The similarity between simulated unfolded images and the correct full-FOV images, evaluated using CW-SSIM (complex wavelet similarity index), was 0.50 ± 0.26, 0.76 ± 0.09, and 0.80 ± 0.10 for the first, second, and third time-point, respectively (numbers closer to 1 indicate more similar images). For the dynamic FOV tests, an average of 11% of the enhancing voxels above the chest wall overlapped (range 0%-40%) due to greater aliasing at early time-points. Despite more voxels overlapping, the CW-SSIM values for the data acquired with dynamic FOVs were 0.64 ± 0.25, 0.93 ± 0.04, and 0.97 ± 0.02 for the first, second, and third time-points, respectively.

CONCLUSIONS

Dynamic FOV imaging allows accelerated bilateral breast DCE-MRI during the early contrast media uptake phase. This method relies on the sparsity of enhancement at the early phases of DCE-MRI of the breast. The results of simulations suggest that dynamic FOV imaging and unfolding produces images that are very close to fully sampled images, and allows temporal resolution as high as 2 s per image.

A g-factor metric for k-t SENSE and k-t PCA based parallel imaging

PURPOSE

To propose and validate a g-factor formalism for k-t SENSE, k-t PCA and related k-t methods for assessing SNR and temporal fidelity.

METHODS

An analytical gxf -factor formulation in the spatiotemporal frequency domain is derived, enabling assessment of noise and depiction fidelity in both the spatial and frequency domain. Using pseudoreplica analysis of cardiac cine data the gxf -factor description is validated and example data are used to analyze the performance of k-t methods for various parameter settings.

RESULTS

Analytical gxf -factor maps were found to agree well with pseudoreplica analysis for 3x, 5x, and 7x k-t SENSE and k-t PCA. While k-t SENSE resulted in lower average gxf values (gx (avg) ) in static regions when compared with k-t PCA, k-t PCA yielded lower gx (avg) values in dynamic regions. Temporal transfer was better preserved with k-t PCA for increasing undersampling factors.

CONCLUSION

The proposed gxf -factor and temporal transfer formalism allows assessing noise performance and temporal depiction fidelity of k-t methods including k-t SENSE and k-t PCA. The framework enables quantitative comparison of different k-t methods relative to frame-by-frame parallel imaging reconstruction.

A g-factor metric for k-t-GRAPPA- and PEAK-GRAPPA-based parallel imaging

PURPOSE

The aim of this work is to derive a theoretical framework for quantitative noise and temporal fidelity analysis of time-resolved k-space-based parallel imaging methods.

THEORY

An analytical formalism of noise distribution is derived extending the existing g-factor formulation for nontime-resolved generalized autocalibrating partially parallel acquisition (GRAPPA) to time-resolved k-space-based methods. The noise analysis considers temporal noise correlations and is further accompanied by a temporal filtering analysis.

METHODS

All methods are derived and presented for k-t-GRAPPA and PEAK-GRAPPA. A sliding window reconstruction and nontime-resolved GRAPPA are taken as a reference. Statistical validation is based on series of pseudoreplica images. The analysis is demonstrated on a short-axis cardiac CINE dataset.

RESULTS

The superior signal-to-noise performance of time-resolved over nontime-resolved parallel imaging methods at the expense of temporal frequency filtering is analytically confirmed. Further, different temporal frequency filter characteristics of k-t-GRAPPA, PEAK-GRAPPA, and sliding window are revealed.

CONCLUSION

The proposed analysis of noise behavior and temporal fidelity establishes a theoretical basis for a quantitative evaluation of time-resolved reconstruction methods. Therefore, the presented theory allows for comparison between time-resolved parallel imaging methods and also nontime-resolved methods. Magn Reson Med 74:125-135, 2015. © 2014 Wiley Periodicals, Inc.

Accelerated fast spin-echo magnetic resonance imaging of the heart using a self-calibrated split-echo approach

Purpose

Design, validation and application of an accelerated fast spin-echo (FSE) variant that uses a split-echo approach for self-calibrated parallel imaging.

Methods

For self-calibrated, split-echo FSE (SCSE-FSE), extra displacement gradients were incorporated into FSE to decompose odd and even echo groups which were independently phase encoded to derive coil sensitivity maps, and to generate undersampled data (reduction factor up to R = 3). Reference and undersampled data were acquired simultaneously. SENSE reconstruction was employed.

Results

The feasibility of SCSE-FSE was demonstrated in phantom studies. Point spread function performance of SCSE-FSE was found to be competitive with traditional FSE variants. The immunity of SCSE-FSE for motion induced mis-registration between reference and undersampled data was shown using a dynamic left ventricular model and cardiac imaging. The applicability of black blood prepared SCSE-FSE for cardiac imaging was demonstrated in healthy volunteers including accelerated multi-slice per breath-hold imaging and accelerated high spatial resolution imaging.

Conclusion

SCSE-FSE obviates the need of external reference scans for SENSE reconstructed parallel imaging with FSE. SCSE-FSE reduces the risk for mis-registration between reference scans and accelerated acquisitions. SCSE-FSE is feasible for imaging of the heart and of large cardiac vessels but also meets the needs of brain, abdominal and liver imaging.

Real-time imaging with radial GRAPPA: Implementation on a heterogeneous architecture for low-latency reconstructions

Combination of non-Cartesian trajectories with parallel MRI permits to attain unmatched acceleration rates when compared to traditional Cartesian MRI during real-time imaging. However, computationally demanding reconstructions of such imaging techniques, such as k-space domain radial generalized auto-calibrating partially parallel acquisitions (radial GRAPPA) and image domain conjugate gradient sensitivity encoding (CG-SENSE), lead to longer reconstruction times and unacceptable latency for online real-time MRI on conventional computational hardware. Though CG-SENSE has been shown to work with low-latency using a general purpose graphics processing unit (GPU), to the best of our knowledge, no such effort has been made for radial GRAPPA. Radial GRAPPA reconstruction, which is robust even with highly undersampled acquisitions, is not iterative, requiring only significant computation during initial calibration while achieving good image quality for low-latency imaging applications. In this work, we present a very fast, low-latency, reconstruction framework based on a heterogeneous system using multi-core CPUs and GPUs. We demonstrate an implementation of radial GRAPPA that permits reconstruction times on par with or faster than acquisition of highly accelerated datasets in both cardiac and dynamic musculoskeletal imaging scenarios. Acquisition and reconstruction times are reported.

High spatial and temporal resolution retrospective cine cardiovascular magnetic resonance from shortened free breathing real-time acquisitions

BACKGROUND

Cine cardiovascular magnetic resonance (CMR) is challenging in patients who cannot perform repeated breath holds. Real-time, free-breathing acquisition is an alternative, but image quality is typically inferior. There is a clinical need for techniques that achieve similar image quality to the segmented cine using a free breathing acquisition. Previously, high quality retrospectively gated cine images have been reconstructed from real-time acquisitions using parallel imaging and motion correction. These methods had limited clinical applicability due to lengthy acquisitions and volumetric measurements obtained with such methods have not previously been evaluated systematically.

METHODS

This study introduces a new retrospective reconstruction scheme for real-time cine imaging which aims to shorten the required acquisition. A real-time acquisition of 16-20s per acquired slice was inputted into a retrospective cine reconstruction algorithm, which employed non-rigid registration to remove respiratory motion and SPIRiT non-linear reconstruction with temporal regularization to fill in missing data. The algorithm was used to reconstruct cine loops with high spatial (1.3-1.8 × 1.8-2.1 mm²) and temporal resolution (retrospectively gated, 30 cardiac phases, temporal resolution 34.3 ± 9.1 ms). Validation was performed in 15 healthy volunteers using two different acquisition resolutions (256 × 144/192 × 128 matrix sizes). For each subject, 9 to 12 short axis and 3 long axis slices were imaged with both segmented and real-time acquisitions. The retrospectively reconstructed real-time cine images were compared to a traditional segmented breath-held acquisition in terms of image quality scores. Image quality scoring was performed by two experts using a scale between 1 and 5 (poor to good). For every subject, LAX and three SAX slices were selected and reviewed in the random order. The reviewers were blinded to the reconstruction approach and acquisition protocols and scores were given to segmented and retrospective cine series. Volumetric measurements of cardiac function were also compared by manually tracing the myocardium for segmented and retrospective cines.

RESULTS

Mean image quality scores were similar for short axis and long axis views for both tested resolutions. Short axis scores were 4.52/4.31 (high/low matrix sizes) for breath-hold vs. 4.54/4.56 for real-time (paired t-test, P = 0.756/0.011). Long axis scores were 4.09/4.37 vs. 3.99/4.29 (P = 0.475/0.463). Mean ejection fraction was 60.8/61.4 for breath-held acquisitions vs. 60.3/60.3 for real-time acquisitions (P = 0.439/0.093). No significant differences were seen in end-diastolic volume (P = 0.460/0.268) but there was a trend towards a small overestimation of end-systolic volume of 2.0/2.5 ml, which did not reach statistical significance (P = 0.052/0.083).

CONCLUSIONS

Real-time free breathing CMR can be used to obtain high quality retrospectively gated cine images in 16-20s per slice. Volumetric measurements and image quality scores were similar in images from breath-held segmented and free breathing, real-time acquisitions. Further speedup of image reconstruction is still needed.

Improved partial k-space reconstruction technique for dynamic myocardial perfusion MRI

The advent of partially parallel k-space data acquisition and reconstruction techniques has resulted in a drastic reduction in acquisition time. In this work, we have shown that the use of a new hybrid technique for reconstruction of partial k-space data can reduce acquisition time without loss of spatiotemporal resolution. The technique is termed hybrid k-t GRAPPA since it is a combination of k-t GRAPPA and an image-k-space keyhole technique. The hybrid method with an acceleration factor 4 generated better results than standard GRAPPA with acceleration factor 2 in preliminary results.

Null space imaging: nonlinear magnetic encoding fields designed complementary to receiver coil sensitivities for improved acceleration in parallel imaging

To increase image acquisition efficiency, we develop alternative gradient encoding strategies designed to provide spatial encoding complementary to the spatial encoding provided by the multiple receiver coil elements in parallel image acquisitions. Intuitively, complementary encoding is achieved when the magnetic field encoding gradients are designed to encode spatial information where receiver spatial encoding is ambiguous, for example, along sensitivity isocontours. Specifically, the method generates a basis set for the null space of the coil sensitivities with the singular value decomposition and calculates encoding fields from the null space vectors. A set of nonlinear gradients is used as projection imaging readout magnetic fields, replacing the conventional linear readout field and phase encoding. Multiple encoding fields are used as projections to capture the null space information, hence the term null space imaging. The method is compared to conventional Cartesian SENSitivity Encoding as evaluated by mean squared error and robustness to noise. Strategies for developments in the area of nonlinear encoding schemes are discussed. The null space imaging approach yields a parallel imaging method that provides high acceleration factors with a limited number of receiver coil array elements through increased time efficiency in spatial encoding.

Automatic and robust calibration of optical detector arrays for biomedical diffuse optical spectroscopy

The design and testing of a new, fully automated, calibration approach is described. The process was used to calibrate an image-guided diffuse optical spectroscopy system with 16 photomultiplier tubes (PMTs), but can be extended to any large array of optical detectors and associated imaging geometry. The design goals were accomplished by developing a routine for robust automated calibration of the multi-detector array within 45 minutes. Our process was able to characterize individual detectors to a median norm of the residuals of 0.03 V for amplitude and 4.4 degrees in phase and achieved less than 5% variation between all the detectors at the 95% confidence interval for equivalent measurements. Repeatability of the calibrated data from the imaging system was found to be within 0.05 V for amplitude and 0.2 degrees for phase, and was used to evaluate tissue-simulating phantoms in two separate imaging geometries. Spectroscopic imaging of total hemoglobin concentration was recovered to within 5% of the true value in both cases. Future work will focus on streamlining the technology for use in a clinical setting with expectations of achieving accurate quantification of suspicious lesions in the breast.

Analytic image concept combined to SENSE reconstruction

OBJECT

Two approaches of reconstructing undersampled partial k-space data, acquired with multiple coils are compared: homodyne detection combined with SENSE (HM_SENSE) and analytic image reconstruction combined with SENSE (AI_SENSE). The latter overcomes limitations of HM_ SENSE by considering aliased images as analytic thus avoiding the need for phase correction required for HM_SENSE.

MATERIALS AND METHODS

In vivo imaging experiments were carried out in male Lewis rats using both gradient echo and spin echo sequences. Accelerated images obtained by using the various reconstruction algorithms were compared to fully sampled reference images both qualitatively and quantitatively.

RESULTS

For the various sampling patterns evaluated, both HM_SENSE and AI_SENSE were found to yield robust image reconstruction with small deviations from the reference image. Even for high acceleration factors AI_SENSE still provided useful results and was found superior compared to HM_SENSE.

CONCLUSION

Combination of partial k-space sampling and parallel image acquisition allows for further acceleration of data acquisition as compared to each method alone. Image reconstruction from undersampled data sets using the AI_SENSE algorithm was found to considerably reduce reconstruction errors and artifacts observed for HM_SENSE reconstruction caused by errors in phase estimation.

TRIO a technique for reconstruction using intensity order: application to undersampled MRI

Long acquisition times are still a limitation for many applications of magnetic resonance imaging (MRI), specially in 3-D and dynamic imaging. Several undersampling reconstruction techniques have been proposed to overcome this problem. These techniques are based on acquiring less samples than specified by the Nyquist criterion and estimating the nonacquired data by using some sort of prior information. Most of these reconstruction methods use prior information based on estimations of the pixel intensities of the images and therefore they are prone to introduce spatial or temporal blurring. Instead of using the pixel intensities, we propose to use information that allows us to sort the pixels of an image from darkest to brightest. The set of order relations which sort the pixels of an image has been called intensity order. The intensity order of an image can be estimated from low-resolution images, adjacent slices in volumetric acquisitions, temporal correlation in dynamic sequences or from prior reconstructions. Our technique for reconstruction using intensity order (TRIO) consists of looking for an image that satisfies the intensity order and minimizes the discrepancy between the acquired and reconstructed data. Results show that TRIO can effectively reconstruct 2-D-cine cardiac MR images (under-sampling factor of 4), estimating correctly the temporal evolution of the objects. Furthermore, TRIO is used as a second stage reconstruction after reconstructing with other techniques, keyhole, sliding window and k-t BLAST, to estimate the order information. In all cases the images are improved by TRIO.

Combining two-dimensional spatially selective RF excitation, parallel imaging, and UNFOLD for accelerated MR thermometry imaging

MR thermometry can be a very challenging application, as good resolution may be needed along spatial, temporal, and temperature axes. Given that the heated foci produced during thermal therapies are typically much smaller than the anatomy being imaged, much of the imaged field-of-view is not actually being heated and may not require temperature monitoring. In this work, many-fold improvements were obtained in terms of temporal resolution and/or 3D spatial coverage by sacrificing some of the in-plane spatial coverage. To do so, three fast-imaging approaches were jointly implemented with a spoiled gradient echo sequence: (1) two-dimensional spatially selective RF excitation, (2) unaliasing by Fourier encoding the overlaps using the temporal dimension (UNFOLD), and (3) parallel imaging. The sequence was tested during experiments with focused ultrasound heating in ex vivo tissue and a tissue-mimicking phantom. Temperature maps were estimated from phase-difference images based on the water proton resonance frequency shift. Results were compared to those obtained from a spoiled gradient echo sequence sequence, using a t-test. Temporal resolution was increased by 24-fold, with temperature uncertainty less than 1°C, while maintaining accurate temperature measurements (mean difference between measurements, as observed in gel = 0.1°C ± 0.6; R = 0.98; P > 0.05).

Retrospectively gated cardiac cine imaging with temporal and spatial acceleration

Parallel imaging methods are routinely used to accelerate the image acquisition process in cardiac cine imaging. The addition of a temporal acceleration method, whereby k-space is sampled differently for different time frames, has been shown in prior work to improve image quality as compared to parallel imaging by itself. However, such temporal acceleration strategies prove difficult to combine with retrospectively gated cine imaging. The only currently published method to feature such combination, by Hansen et al. [Magn Reson Med 55 (2006) 85-91] tends to be associated with prohibitively long reconstruction times. The goal of the present work was to develop a retrospectively gated cardiac cine method that features both parallel imaging and temporal acceleration, capable of achieving significant acceleration factors on commonly available hardware and associated with reconstruction times short enough for practical use in a clinical context. Seven cardiac patients and a healthy volunteer were recruited and imaged, with acceleration factors of 3.5 or 4.5, using an eight-channel product cardiac array on a 1.5-T system. The prescribed FOV value proved slightly too small in three patients, and one of the patients had a bigemini condition. Despite these additional challenges, good-quality results were obtained for all slices and all patients, with a reconstruction time of 0.98±0.07 s per frame, or about 20 s for a 20-frame slice, using a single processor on a single PC. As compared to using parallel imaging by itself, the addition of a temporal acceleration strategy provided much resistance to artifacts.

Temporal filtering effects in dynamic parallel MRI

Autocalibrated parallel MRI methods such as TSENSE or k-t SENSE have been presented for dynamic imaging studies as they are able to provide images with high temporal resolution. One key element of these techniques is the temporal averaging of the undersampled raw data to obtain an unaliased image. This image represents the temporal average (also known as direct current, DC) and is used to derive the reconstruction parameters. In this work, we show that aliasing artifacts can be introduced in the DC signal obtained from the undersampled raw data. These artifacts lead to undesired temporal filtering effects when the DC signal is used for coil sensitivity calibration or when the DC signal is subtracted from the raw data. It is demonstrated that the temporal filtering effects can be reduced significantly by filtering the DC signal.

Patient-adaptive reconstruction and acquisition in dynamic imaging with sensitivity encoding (PARADISE)

MRI of the human heart without explicit cardiac synchronization promises to extend the applicability of cardiac MR to a larger patient population and potentially expand its diagnostic capabilities. However, conventional nongated imaging techniques typically suffer from low image quality or inadequate spatio-temporal resolution and fidelity. Patient-Adaptive Reconstruction and Acquisition in Dynamic Imaging with Sensitivity Encoding (PARADISE) is a highly accelerated nongated dynamic imaging method that enables artifact-free imaging with high spatio-temporal resolutions by utilizing novel computational techniques to optimize the imaging process. In addition to using parallel imaging, the method gains acceleration from a physiologically driven spatio-temporal support model; hence, it is doubly accelerated. The support model is patient adaptive, i.e., its geometry depends on dynamics of the imaged slice, e.g., subject's heart rate and heart location within the slice. The proposed method is also doubly adaptive as it adapts both the acquisition and reconstruction schemes. Based on the theory of time-sequential sampling, the proposed framework explicitly accounts for speed limitations of gradient encoding and provides performance guarantees on achievable image quality. The presented in-vivo results demonstrate the effectiveness and feasibility of the PARADISE method for high-resolution nongated cardiac MRI during short breath-hold.

Breathheld autocalibrated phase-contrast imaging

PURPOSE

To compare generalized autocalibrating partially parallel acquisitions (GRAPPA), modified sensitivity encoding (mSENSE), and SENSE in phase-contrast magnetic resonance imaging (PC-MRI) applications.

MATERIALS AND METHODS

Aliasing of the torso can occur in PC-MRI applications. If the data are further undersampled for parallel imaging, SENSE can be problematic in correctly unaliasing signals due to coil sensitivity maps that do not match that of the aliased volume. Here, a method for estimating coil sensitivities in flow applications is described. Normal volunteers (n = 5) were scanned on a 1.5 T MRI scanner and underwent PC-MRI scans using GRAPPA, mSENSE, SENSE, and conventional PC-MRI acquisitions. Peak velocity and flow through the aorta and pulmonary artery were evaluated.

RESULTS

Bland-Altman statistics for flow in the aorta and pulmonary artery acquired with mSENSE and GRAPPA methods (R = 2 and R = 3 cases) have comparable mean differences to flow acquired with conventional PC-MRI. GRAPPA and mSENSE PC-MRI have more robust measurements than SENSE when there is aliasing artifact caused by insufficient coil sensitivity maps. For peak velocity, there are no considerable differences among the mSENSE, GRAPPA, and SENSE reconstructions and are comparable to conventional PC-MRI.

CONCLUSION

Flow measurements of images reconstructed with autocalibration techniques have comparable agreement with conventional PC-MRI and provide robust measurements in the presence of wraparound.

Effect of improving spatial or temporal resolution on image quality and quantitative perfusion assessment with k-t SENSE acceleration in first-pass CMR myocardial perfusion imaging

k-t Sensitivity-encoded (k-t SENSE) acceleration has been used to improve spatial resolution, temporal resolution, and slice coverage in first-pass cardiac magnetic resonance myocardial perfusion imaging. This study compares the effect of investing the speed-up afforded by k-t SENSE acceleration in spatial or temporal resolution. Ten healthy volunteers underwent adenosine stress myocardial perfusion imaging using four saturation-recovery gradient echo perfusion sequences: a reference sequence accelerated by sensitivity encoding (SENSE), and three k-t SENSE-accelerated sequences with higher spatial resolution ("k-t High"), shorter acquisition window ("k-t Fast"), or a shared increase in both parameters ("k-t Hybrid") relative to the reference. Dark-rim artifacts and image quality were analyzed. Semiquantitative myocardial perfusion reserve index (MPRI) and Fermi-derived quantitative MPR were also calculated. The k-t Hybrid sequence produced highest image quality scores at rest (P = 0.015). Rim artifact thickness and extent were lowest using k-t High and k-t Hybrid sequences (P < 0.001). There were no significant differences in MPRI and MPR values derived by each sequence. Maximizing spatial resolution by k-t SENSE acceleration produces the greatest reduction in dark rim artifact. There is good agreement between k-t SENSE and standard acquisition methods for semiquantitative and fully quantitative myocardial perfusion analysis.

A 2D MTF approach to evaluate and guide dynamic imaging developments

As the number and complexity of partially sampled dynamic imaging methods continue to increase, reliable strategies to evaluate performance may prove most useful. In the present work, an analytical framework to evaluate given reconstruction methods is presented. A perturbation algorithm allows the proposed evaluation scheme to perform robustly without requiring knowledge about the inner workings of the method being evaluated. A main output of the evaluation process consists of a two-dimensional modulation transfer function, an easy-to-interpret visual rendering of a method's ability to capture all combinations of spatial and temporal frequencies. Approaches to evaluate noise properties and artifact content at all spatial and temporal frequencies are also proposed. One fully sampled phantom and three fully sampled cardiac cine datasets were subsampled (R = 4 and 8) and reconstructed with the different methods tested here. A hybrid method, which combines the main advantageous features observed in our assessments, was proposed and tested in a cardiac cine application, with acceleration factors of 3.5 and 6.3 (skip factors of 4 and 8, respectively). This approach combines features from methods such as k-t sensitivity encoding, unaliasing by Fourier encoding the overlaps in the temporal dimension-sensitivity encoding, generalized autocalibrating partially parallel acquisition, sensitivity profiles from an array of coils for encoding and reconstruction in parallel, self, hybrid referencing with unaliasing by Fourier encoding the overlaps in the temporal dimension and generalized autocalibrating partially parallel acquisition, and generalized autocalibrating partially parallel acquisition-enhanced sensitivity maps for sensitivity encoding reconstructions.

Real-time cardiac MRI using prior spatial-spectral information

Cardiac MRI performed while the patient is breathing is typically achieved using non-real-time techniques such as ECG triggering with respiratory gating; however, modern dynamic imaging techniques are beginning to enable this type of imaging in real-time. One of these dynamic imaging techniques is based on forming a Partially Separable Function (PSF) model of the data, but the model fitting process is known to be sensitive even when truncated SVD regularization is used. As a result, physiologically meaningless artifacts can appear in the dynamic images when the total number of measurements is limited. To address this issue, the dynamic imaging problem is formulated as a generalized Tikhonov regularization problem with the PSF model as a component of the forward data model, and a penalty function is used to introduce spatial-spectral prior information. This new method both reduces data acquisition requirements and improves stability relative to the original PSF based method when applied to cardiac MRI.

Accelerating SENSE using compressed sensing

Both parallel MRI and compressed sensing (CS) are emerging techniques to accelerate conventional MRI by reducing the number of acquired data. The combination of parallel MRI and CS for further acceleration is of great interest. In this paper, we propose a novel method to combine sensitivity encoding (SENSE), one of the standard methods for parallel MRI, and compressed sensing for rapid MR imaging (SparseMRI), a recently proposed method for applying CS in MR imaging with Cartesian trajectories. The proposed method, named CS-SENSE, sequentially reconstructs a set of aliased reduced-field-of-view images in each channel using SparseMRI and then reconstructs the final image from the aliased images using Cartesian SENSE. The results from simulations and phantom and in vivo experiments demonstrate that CS-SENSE can achieve a reduction factor higher than those achieved by SparseMRI and SENSE individually and outperform the existing method that combines parallel MRI and CS.

Accelerated focused ultrasound imaging

One of the most basic trade-offs in ultrasound imaging involves frame rate, depth, and number of lines. Achieving good spatial resolution and coverage requires a large number of lines, leading to decreases in frame rate. An even more serious imaging challenge occurs with imaging modes involving spatial compounding and 3-D/4-D imaging, which are severely limited by the slow speed of sound in tissue. The present work can overcome these traditional limitations, making ultrasound imaging many-fold faster. By emitting several beams at once, and by separating the resulting overlapped signals through spatial and temporal processing, spatial resolution and/or coverage can be increased by many-fold while leaving frame rates unaffected. The proposed approach can also be extended to imaging strategies that do not involve transmit beamforming, such as synthetic aperture imaging. Simulated and experimental results are presented where imaging speed is improved by up to 32-fold, with little impact on image quality. Object complexity has little impact on the method's performance, and data from biological systems can readily be handled. The present work may open the door to novel multiplexed and/or multidimensional protocols considered impractical today.

k-t PCA: temporally constrained k-t BLAST reconstruction using principal component analysis

The k-t broad-use linear acquisition speed-up technique (BLAST) has become widespread for reducing image acquisition time in dynamic MRI. In its basic form k-t BLAST speeds up the data acquisition by undersampling k-space over time (referred to as k-t space). The resulting aliasing is resolved in the Fourier reciprocal x-f space (x = spatial position, f = temporal frequency) using an adaptive filter derived from a low-resolution estimate of the signal covariance. However, this filtering process tends to increase the reconstruction error or lower the achievable acceleration factor. This is problematic in applications exhibiting a broad range of temporal frequencies such as free-breathing myocardial perfusion imaging. We show that temporal basis functions calculated by subjecting the training data to principal component analysis (PCA) can be used to constrain the reconstruction such that the temporal resolution is improved. The presented method is called k-t PCA.

Accelerated CMR using zonal, parallel and prior knowledge driven imaging methods

Accelerated imaging is highly relevant for many CMR applications as competing constraints with respect to spatiotemporal resolution and tolerable scan times are frequently posed. Three approaches, all involving data undersampling to increase scan efficiencies, are discussed in this review. Zonal imaging can be considered a niche but nevertheless has found application in coronary imaging and CMR flow measurements. Current work on parallel-transmit systems is expected to revive the interest in zonal imaging techniques. The second and main approach to speeding up CMR sequences has been parallel imaging. A wide range of CMR applications has benefited from parallel imaging with reduction factors of two to three routinely applied for functional assessment, perfusion, viability and coronary imaging. Large coil arrays, as are becoming increasingly available, are expected to support reduction factors greater than three to four in particular in combination with 3D imaging protocols. Despite these prospects, theoretical work has indicated fundamental limits of coil encoding at clinically available magnetic field strengths. In that respect, alternative approaches exploiting prior knowledge about the object being imaged as such or jointly with parallel imaging have attracted considerable attention. Five to eight-fold scan accelerations in cine and dynamic CMR applications have been reported and image quality has been found to be favorable relative to using parallel imaging alone.With all acceleration techniques, careful consideration of the limits and the trade-off between acceleration and occurrence of artifacts that may arise if these limits are breached is required. In parallel imaging the spatially varying noise has to be considered when measuring contrast- and signal-to-noise ratios. Also, temporal fidelity in images reconstructed with prior knowledge driven methods has to be studied carefully.

On the complimentarity of SENSE and GRAPPA in parallel MR imaging

Two image reconstruction methods currently dominate parallel MR imaging: SENSE and GRAPPA. While both seek to reconstruct images from subsampled multi-channel MRI data, there exist fundamental differences between the two. In particular, SENSE reconstructs an image of the excited spin-density directly whereas GRAPPA reconstructs estimates of the fully sampled raw coil data and then combines them to obtain an image. In this work we show that these differences can be exploited such that each method can compliment the other. In the case of SENSE, which requires an estimate of the coil sensitivity map before reconstruction, one can use GRAPPA to improve the coil sensitivity estimates. Alternatively, using coil sensitivity estimates and the SENSE reconstruction equations, one can improve the GRAPPA reconstruction parameter estimation. Together, these approaches can provide higher image quality than either method alone.

Highly accelerated cardiovascular MR imaging using many channel technology: concepts and clinical applications

Cardiovascular magnetic resonance imaging (CVMRI) is of proven clinical value in the non-invasive imaging of cardiovascular diseases. CVMRI requires rapid image acquisition, but acquisition speed is fundamentally limited in conventional MRI. Parallel imaging provides a means for increasing acquisition speed and efficiency. However, signal-to-noise (SNR) limitations and the limited number of receiver channels available on most MR systems have in the past imposed practical constraints, which dictated the use of moderate accelerations in CVMRI. High levels of acceleration, which were unattainable previously, have become possible with many-receiver MR systems and many-element, cardiac-optimized RF-coil arrays. The resulting imaging speed improvements can be exploited in a number of ways, ranging from enhancement of spatial and temporal resolution to efficient whole heart coverage to streamlining of CVMRI work flow. In this review, examples of these strategies are provided, following an outline of the fundamentals of the highly accelerated imaging approaches employed in CVMRI. Topics discussed include basic principles of parallel imaging; key requirements for MR systems and RF-coil design; practical considerations of SNR management, supported by multi-dimensional accelerations, 3D noise averaging and high field imaging; highly accelerated clinical state-of-the art cardiovascular imaging applications spanning the range from SNR-rich to SNR-limited; and current trends and future directions.

Joint image reconstruction and sensitivity estimation in SENSE (JSENSE)

Parallel magnetic resonance imaging (pMRI) using multichannel receiver coils has emerged as an effective tool to reduce imaging time in various applications. However, the issue of accurate estimation of coil sensitivities has not been fully addressed, which limits the level of speed enhancement achievable with the technology. The self-calibrating (SC) technique for sensitivity extraction has been well accepted, especially for dynamic imaging, and complements the common calibration technique that uses a separate scan. However, the existing method to extract the sensitivity information from the SC data is not accurate enough when the number of data is small, and thus erroneous sensitivities affect the reconstruction quality when they are directly applied to the reconstruction equation. This paper considers this problem of error propagation in the sequential procedure of sensitivity estimation followed by image reconstruction in existing methods, such as sensitivity encoding (SENSE) and simultaneous acquisition of spatial harmonics (SMASH), and reformulates the image reconstruction problem as a joint estimation of the coil sensitivities and the desired image, which is solved by an iterative optimization algorithm. The proposed method was tested on various data sets. The results from a set of in vivo data are shown to demonstrate the effectiveness of the proposed method, especially when a rather large net acceleration factor is used.

High temporal resolution breathheld 3D FIESTA CINE imaging: validation of ventricular function in patients with chronic myocardial infarction

PURPOSE

To develop a gated single-breathhold, high temporal resolution three-dimensional (3D) CINE imaging technique and to evaluate its accuracy in volumetric and functional quantification in patients with chronic myocardial infarction.

MATERIALS AND METHODS

A 3D CINE steady-state free precession (SSFP) pulse sequence was developed incorporating variable temporal sampling of the low and high spatial frequency k-space data to reduce breathhold time and parallel imaging to increase temporal resolution. Reconstruction with retrospective interpolation enabled complete R-R interval coverage. Feasibility was assessed in eight patients with chronic myocardial infarction and ventricular functional values were compared to those of a 2D CINE acquisition.

RESULTS

There was no significant difference between the 3D CINE and 2D CINE for end-diastolic volume (168 +/- 73 vs. 177 +/- 59 mL, respectively; P < 0.27), end-systolic volume (81 +/- 62 vs. 79 +/- 53 mL; P < 0.81), and ejection fraction (EF) measurements (55 +/- 14% vs. 58 +/- 14%; P < 0.14). The mean difference in EF was less than 2.5%. A wall motion assessment indicated a good agreement, with a weighted kappa value of 0.62.

CONCLUSION

High temporal resolution 3D CINE SSFP imaging of the whole heart can be obtained in a single breathhold and yield ventricular function measurements similar to 2D CINE methods.

Improving k-t SENSE by adaptive regularization

The recently proposed method known as k-t sensitivity encoding (SENSE) has emerged as an effective means of improving imaging speed for several dynamic imaging applications. However, k-t SENSE uses temporally averaged data as a regularization term for image reconstruction. This may not only compromise temporal resolution, it may also make some of the temporal frequency components irrecoverable. To address that issue, we present a new method called spatiotemporal domain-based unaliasing employing sensitivity encoding and adaptive regularization (SPEAR). Specifically, SPEAR provides an improvement over k-t SENSE by generating adaptive regularization images. It also uses a variable-density (VD), sequentially interleaved k-t space sampling pattern with reference frames for data acquisition. Simulations based on experimental data were performed to compare SPEAR, k-t SENSE, and several other related methods, and the results showed that SPEAR can provide higher temporal resolution with significantly reduced image artifacts. Ungated 3D cardiac imaging experiments were also carried out to test the effectiveness of SPEAR, and real-time 3D short-axis images of the human heart were produced at 5.5 frames/s temporal resolution and 2.4 x 1.2 x 8 mm3 spatial resolution with eight slices.

Sensitivity-encoded (SENSE) proton echo-planar spectroscopic imaging (PEPSI) in the human brain

Magnetic resonance spectroscopic imaging (MRSI) provides spatially resolved metabolite information that is invaluable for both neuroscience studies and clinical applications. However, lengthy data acquisition times, which are a result of time-consuming phase encoding, represent a major challenge for MRSI. Fast MRSI pulse sequences that use echo-planar readout gradients, such as proton echo-planar spectroscopic imaging (PEPSI), are capable of fast spectral-spatial encoding and thus enable acceleration of image acquisition times. Combining PEPSI with recent advances in parallel MRI utilizing RF coil arrays can further accelerate MRSI data acquisition. Here we investigate the feasibility of ultrafast spectroscopic imaging at high field (3T and 4T) by combining PEPSI with sensitivity-encoded (SENSE) MRI using eight-channel head coil arrays. We show that the acquisition of single-average SENSE-PEPSI data at a short TE (15 ms) can be accelerated to 32 s or less, depending on the field strength, to obtain metabolic images of choline (Cho), creatine (Cre), N-acetyl-aspartate (NAA), and J-coupled metabolites (e.g., glutamate (Glu) and inositol (Ino)) with acceptable spectral quality and localization. The experimentally measured reductions in signal-to-noise ratio (SNR) and Cramer-Rao lower bounds (CRLBs) of metabolite resonances were well explained by both the g-factor and reduced measurement times. Thus, this technology is a promising means of reducing the scan times of 3D acquisitions and time-resolved 2D measurements.

A new strategy for respiration compensation, applied toward 3D free-breathing cardiac MRI

In thorax and abdomen imaging, image quality may be affected by breathing motion. Cardiac MR images are typically obtained while the patient holds his or her breath, to avoid respiration-related artifacts. Although useful, breath-holding imposes constraints on scan duration, which in turn limits the achievable resolution and SNR. Longer scan times would be required to improve image quality, and effective strategies are needed to compensate for respiratory motion. A novel approach at respiratory compensation, targeted toward 3D free-breathing cardiac MRI, is presented here. The method aims at suppressing the negative effects of respiratory-induced cardiac motion while capturing the heart's beating motion. The method is designed so that the acquired data can be reconstructed in two different ways: First, a time series of images is reconstructed to quantify and correct for respiratory motion. Then, the corrected data are reconstructed a final time into a cardiac-phase series of images to capture the heart's beating motion. The method was implemented, and initial results are presented. A cardiac-phase series of 3D images, covering the entire heart, was obtained for two free-breathing volunteers. The present method may prove especially useful in situations where breath-holding is not an option, for example, for very sick, mentally impaired or infant patients.

Extension of the UNFOLD method to include free breathing

Unaliasing by Fourier-encoding the overlaps using the temporal dimension (UNFOLD) is a method to reduce the data acquisition burden in dynamic MRI. The method works by forcing aliased signals to behave in specific ways through time, so that these unwanted signals can be detected and removed. Unexpected events in time, such as displacements caused by breathing, have the potential to disturb the temporal strategy and may affect UNFOLD's ability to suppress aliasing artifacts. This work presents an extension of the UNFOLD method to accommodate temporal encoding disruptions. While the main type of disruption considered here comes from respiratory motion, other types of disruption can be envisioned, such as departures from the usual UNFOLD k-space sampling scheme. This extended version of UNFOLD was incorporated into UNFOLD-sensitivity encoding (UNFOLD-SENSE), and should also be applicable to closely related methods such as temporal SENSE (TSENSE), k-t Broaduse Linear Acquisition Speed up Technique (k-t BLAST), and k-t SENSE. Five patients were imaged with a modified version of a myocardial-perfusion sequence, and UNFOLD was used either alone or in conjunction with SENSE to obtain an acceleration of 2.0 (in three patients) or 3.0 (in two patients). In both cases this extended version of UNFOLD was able to suppress artifacts caused by the presence of breathing motion.

Parallel imaging in cardiovascular MRI: methods and applications

Cardiovascular MR imaging (CVMR) has become a valuable modality for the non-invasive detection and characterization of cardiovascular diseases. CVMR requires high imaging speed and efficiency, which is fundamentally limited in conventional cardiovascular MRI studies. With the introduction of parallel imaging, alternative means for increasing acquisition speed beyond these limits have become available. In parallel imaging some image data are acquired simultaneously, using RF detector coil sensitivities to encode simultaneous spatial information that complements the information gleaned from sequential application of magnetic field gradients. The resulting improvements in imaging speed can be used in various ways, including shortening long examinations, improving spatial resolution and/or anatomic coverage, improving temporal resolution, enhancing image quality, overcoming physiological constraints, detecting and correcting for physiologic motion, and streamlining work flow. Examples of each of these strategies will be provided in this review. First, basic principles and key concepts of parallel MR are described. Second, practical considerations such as coil array design, coil sensitivity calibrations, customized pulse sequences and tailored imaging parameters are outlined. Next, cardiovascular applications of parallel MR are reviewed, ranging from cardiac anatomical and functional assessment to myocardial perfusion and viability to MR angiography of the coronary arteries and the large vessels. Finally, current trends and future directions in parallel CVMR are considered.

k-t GRAPPA: a k-space implementation for dynamic MRI with high reduction factor

A novel technique called "k-t GRAPPA" is introduced for the acceleration of dynamic magnetic resonance imaging. Dynamic magnetic resonance images have significant signal correlations in k-space and time dimension. Hence, it is feasible to acquire only a reduced amount of data and recover the missing portion afterward. Generalized autocalibrating partially parallel acquisitions (GRAPPA), as an important parallel imaging technique, linearly interpolates the missing data in k-space. In this work, it is shown that the idea of GRAPPA can also be applied in k-t space to take advantage of the correlations and interpolate the missing data in k-t space. For this method, no training data, filters, additional parameters, or sensitivity maps are necessary, and it is applicable for either single or multiple receiver coils. The signal correlation is locally derived from the acquired data. In this work, the k-t GRAPPA technique is compared with our implementation of GRAPPA, TGRAPPA, and sliding window reconstructions, as described in Methods. The experimental results manifest that k-t GRAPPA generates high spatial resolution reconstruction without significant loss of temporal resolution when the reduction factor is as high as 4. When the reduction factor becomes higher, there might be a noticeable loss of temporal resolution since k-t GRAPPA uses temporal interpolation. Images reconstructed using k-t GRAPPA have less residue/folding artifacts than those reconstructed by sliding window, much less noise than those reconstructed by GRAPPA, and wider temporal bandwidth than those reconstructed by GRAPPA with residual k-space. k-t GRAPPA is applicable to a wide range of dynamic imaging applications and is not limited to imaging parts with quasi-periodic motion. Since only local information is used for reconstruction, k-t GRAPPA is also preferred for applications requiring real time reconstruction, such as monitoring interventional MRI.

x-f choice: Reconstruction of undersampled dynamic MRI by data-driven alias rejection applied to contrast-enhanced angiography

A technique for reconstructing dynamic undersampled MRI data, termed "x-f choice," was developed and applied to dynamic contrast-enhanced MR angiography (DCE-MRA). Regular undersampling in k-t space (a hybrid of k-space and time) creates aliasing in the conjugate x-f space that must be resolved. When regions in the object containing fast dynamic change are sparse, as in DCE-MRA, signal overlap caused by aliasing is often much less than the undersample factor would imply. x-f Choice reconstruction identifies overlapping signals using a model of the full non-aliased x-f space that is automatically generated from the undersampled data, and applies parallel imaging (PI) to separate them. No extra reference scans are required to generate either the model or the coil sensitivity maps. At each location in the reconstructed images, g-factor noise amplification is compared with predicted reconstruction errors to obtain an optimized solution. Acceleration factors greater than the number of receiver coils are possible, but are limited by the sparseness of the dynamic content and the signal-to-noise ratio (SNR) (in DCE-MRA the latter is dominant). Temporal fidelity was validated for up to a factor 10 speed-up using retrospectively undersampled data from a six-coil array. The method was tested on volunteers using fivefold prospective undersampling.

Functional MRI using regularized parallel imaging acquisition

Parallel MRI techniques reconstruct full-FOV images from undersampled k-space data by using the uncorrelated information from RF array coil elements. One disadvantage of parallel MRI is that the image signal-to-noise ratio (SNR) is degraded because of the reduced data samples and the spatially correlated nature of multiple RF receivers. Regularization has been proposed to mitigate the SNR loss originating due to the latter reason. Since it is necessary to utilize static prior to regularization, the dynamic contrast-to-noise ratio (CNR) in parallel MRI will be affected. In this paper we investigate the CNR of regularized sensitivity encoding (SENSE) acquisitions. We propose to implement regularized parallel MRI acquisitions in functional MRI (fMRI) experiments by incorporating the prior from combined segmented echo-planar imaging (EPI) acquisition into SENSE reconstructions. We investigated the impact of regularization on the CNR by performing parametric simulations at various BOLD contrasts, acceleration rates, and sizes of the active brain areas. As quantified by receiver operating characteristic (ROC) analysis, the simulations suggest that the detection power of SENSE fMRI can be improved by regularized reconstructions, compared to unregularized reconstructions. Human motor and visual fMRI data acquired at different field strengths and array coils also demonstrate that regularized SENSE improves the detection of functionally active brain regions.